In many instances it would be desirable to have a database that captures the geometry of a train track, as well as the geo-locations of various features (e.g., grade crossings, mileposts, signals, platforms, switches, spurs, etc.) along the train track. For example, such a database would be useful in train motion and path-taken navigation algorithms, predictive braking algorithms, and locomotive fuel management algorithms.
If a track database become very large, however, a variety of issues can mitigate the database's utility. For example, a large database typically requires greater effort to build and maintain, is more susceptible to data errors, and has less portability. Moreover, high data-rate links might be required to transmit data to locomotives if the track database is too large.